Design, develop, integrate and optimise scalable AI/ML solutions, including LLM-based applications, Retrieval-Augmented Generation (RAG), agentic AI and multimodal AI solutions to address Home Team use cases.
Design and develop resource scheduling systems for machine learning platforms, supporting model training, evaluation, and inference workloads across domains such as NLP, Computer Vision (CV), and Speech.
Responsible for the overall architecture design and implementation of model inference services, building a high-performance, highly available, and scalable enterprise-level inference system for large-parameter, high-complexity AI models, overcoming various architectural challenges in the implementation of complex model inference, and supporting the efficient launch of models across all business scenarios.
Build and maintain an automated LLM evaluation pipeline covering multiple dimensions, including general capabilities, Agent capabilities, and persona/role-playing. The pipeline should support one-click evaluation, historical result comparison, and regression testing.
Build and maintain an automated LLM evaluation pipeline covering multiple dimensions, including general capabilities, Agent capabilities, and persona/role-playing. The pipeline should support one-click evaluation, historical result comparison, and regression testing.
Build and maintain an automated LLM evaluation pipeline covering multiple dimensions, including general capabilities, Agent capabilities, and persona/role-playing. The pipeline should support one-click evaluation, historical result comparison, and regression testing.
Support and enhance training content such as workbooks, job aids, posters, campaign materials, and presentation decks aligned with HTQA standards and brand guidelines.
Technical Strategy: Define the AIGC roadmap across pre-training and post-training workstreams — from distributed training infrastructure to alignment, evaluation, and inference deployment.
Curating progressive, impactful developmental opportunities to level up the workforce capabilities of the Whole-of-Government Service Delivery function and partnering various stakeholders to implement them
For large-model training and evaluation scenarios, participate in formulating and continuously iterating data quality standards, judgment rules, and acceptance criteria.
For large-model training and evaluation scenarios, participate in formulating and continuously iterating data quality standards, judgment rules, and acceptance criteria.
Area A — Scientific ML & PINNs Methodology Origination: Originate and advance PINNs methodology — design physics-constrained loss function architectures, validate digital twin ML components against domain physics (with storage domain expert), and deliver validated prototypes with complete technical documentation to implement. As the sole PINNs methodology originator on the team — this capability cannot be delegated or substituted.
Participate in the R&D of TikTok Shop's Southeast Asia business, including end-to-end full-stack delivery spanning front-end pages, server-side APIs, data pipelines, and launch verification;